In 1880 George Mortimer Pullman, who had made one fortune raising Chicago’s buildings out of the mud on jackscrews and a much larger one manufacturing the sleeping cars that carried Americans across their continent, began buying four thousand acres of prairie beside Lake Calumet, about fourteen miles south of the city. On it he built a factory and, around the factory, a town. It had brick houses with indoor plumbing and gas, paved streets, sewers, garbage collection, a hotel, an arcade containing shops and a library and a theatre, a market hall, a church, parks, playing fields, and an artificial lake. In an era when Chicago’s working districts were a byword for filth and disease, this was not a gesture. It was the best housing available to industrial workers anywhere in the United States, and in 1896 an international exposition in Prague named it the world’s most perfect town.
Fourteen years after the first residents moved in, that town was the origin of the largest labour conflict in nineteenth-century America: a strike that became a national boycott, halted rail traffic from Ohio to California across twenty-seven states, drew in something approaching half the United States Army, and left around thirty people dead. Four years after that, the Illinois Supreme Court ordered the company to sell the town, on the grounds that its charter said nothing about owning one and that company towns were contrary to public policy. The usual explanation is that Pullman was a hypocrite whose model town was a sham, and the record does not support that; the houses were genuinely good and are still standing and still lived in. What went wrong was structural, and it was built into the arrangement from the first survey stake, which puts this case somewhere apart from most of the record of planned communities and alongside the great industrial paternalisms such as Ford’s plantation town on the Amazon.
The Man Who Raised Chicago
Pullman’s career before the town explains a good deal about how he approached it. He came to Chicago in his twenties and made his reputation on an audacious piece of engineering: lifting entire multi-storey brick buildings, including the six-storey Tremont House, several feet out of the mud on which the city was sinking, using hundreds of jackscrews turned in unison, without cracking the masonry or interrupting the businesses inside. It is a very particular kind of achievement, and it tells you what sort of mind he had. Coordinate enough small forces precisely enough and you can move something everyone regards as immovable.
The sleeping car business is worth understanding because it shaped how Pullman thought about labour. His company did not merely build cars and sell them; it built them, retained ownership, and leased them to the railroads complete with staff, so that Pullman was simultaneously a manufacturer, a lessor, and a service employer across the entire national network. That vertically integrated model was enormously profitable and it required an unusual degree of central control over standards, staffing, and conduct, since the product being sold was a uniform experience delivered by employees the railroads did not manage. A man accustomed to running a service business at national scale is accustomed to specifying how his people behave, and the manufacture of physical goods to consistent standards, which depends on exactly that kind of process discipline, as any modern industrial supply chain demonstrates.
He then applied the same disposition to rail travel, designing a sleeping car of unprecedented luxury and, in a promotional stroke of considerable brilliance, arranging for his prototype to be attached to Abraham Lincoln’s funeral train in 1865. The Pullman car became the standard of American long-distance travel, the company grew into one of the largest manufacturers in the country, and Pullman became one of the grandees of Chicago society. The engineering achievement was real and belongs in the record of what people manage to build, as does the systems thinking behind it, of a kind that recurs throughout the catalogue of ambitious technical undertakings. He was a genuinely gifted engineer. He approached a town the same way.
Building Pullman
The design work went to Solon Spencer Beman, an architect in his twenties, and to the landscape architect Nathan F. Barrett, and they produced something coherent and genuinely handsome. Brick row houses in varied styles lined tree-planted streets. The Hotel Florence, named for Pullman’s daughter, anchored one end; the Arcade Building, containing shops, a bank, a theatre, and a library stocked with several thousand volumes, anchored the civic centre. The Greenstone Church stood on the square. Water, gas, and sewerage were laid throughout, waste was collected, and the sewage was pumped to a company farm and used as fertiliser, which was advanced sanitary engineering for 1881 and reflected serious thinking about water and waste as systems to be managed.
The intention behind the design was explicit and it was not sentimental. Pullman believed that a workforce housed in clean, orderly, attractive surroundings, removed from the saloons and agitators of the city, would be healthier, steadier, more productive, and less inclined to organise, and he said so. The town was a labour-management instrument justified in the language of uplift, which is not the same as insincerity; he appears to have believed both halves. Environments do shape behaviour, and the proposition that a decent physical setting produces better outcomes is well founded across the study of how organisms respond to their surroundings. The difficulty was never the premise. It was what he attached to it.
The quality was not uniform and it is worth saying so. Housing was allocated by rank, with detached houses of eight or nine rooms near the works for executives, row houses for foremen, smaller quarters for skilled workers, and two-room apartments for the unskilled, while dwellings on the outskirts were more cheaply built and some lacked plumbing altogether. Rents ran perhaps a quarter above comparable accommodation in the surrounding district. But the central district was excellent, the mortality statistics were far better than Chicago’s, and visitors were taken on guided tours and came away impressed, which they were meant to do. Pullman had built something that worked and that photographed beautifully. That was never the problem.
Six Per Cent
Here is the covenant that determined everything else, and Pullman was entirely open about it. The town was not philanthropy. It was an investment, and it was required to return a profit on the capital sunk into it, a figure usually cited at around six per cent and sometimes seven. The National Park Service’s own account of the site records the arrangement plainly: the company owned every building and set rents to ensure a return on its investment in building the town, as the federal history of the site states.
It is worth appreciating how unusual this was among model towns of the period, because a number of industrialists built worker housing and most did not insist that it turn a profit. Some treated it as a cost of doing business in a remote location, some as philanthropy, some as an investment expected merely to break even. Pullman’s insistence that the town yield a commercial return on capital, comparable to what the money would earn elsewhere, was a deliberate and much-publicised principle rather than an accounting accident: he argued that charity was demoralising and that an arrangement which paid its way was more honest and more durable than one that depended on a benefactor’s continuing generosity. There is a real argument in that. It simply commits you to collecting the rent in a depression.
That requirement is not a detail; it is the governing constraint from which the town’s peculiarities follow by deduction. If the housing must return six per cent, then rents are set by the capital account rather than by what tenants earn, and they cannot fall when wages fall without breaching the covenant. If the housing must return six per cent, then nobody can be permitted to buy a house, because selling the asset liquidates the income stream, and so every resident in Pullman was a tenant at will, on a lease terminable at ten days’ notice. If the housing must return six per cent, then no independent landlord can be allowed to build nearby and undercut, and no independent shopkeeper can be allowed to compete with the company’s arcade rents. And if all of that is to hold, then the town cannot be permitted to govern itself, because a municipality of tenants would very quickly vote on rents. The financial requirement and the political arrangement are the same requirement stated twice, which is the sort of thing that becomes visible only when the accounts and the constitution are read together, as every serious examination of how ownership structures shape conduct finds. It had to pay. Everything else follows from that.
What the Company Owned
The contrast with communities that governed themselves is instructive and was available at the time. Zoar and Oneida owned their property collectively and their members were the owners; the self-organising settlements that have lasted, such as the free town in Copenhagen, are governed by the people who live in them. Pullman inverted that completely. The residents owned nothing, decided nothing, and could be removed at ten days’ notice, while the entity that owned and decided everything did not live there. That is not a variation on communal ownership; it is its precise opposite, and the fact that both arrangements get filed under model communities has confused the comparison ever since.
Enumerate the holdings and the shape of the thing becomes plain. The Pullman Palace Car Company owned the factory where a man worked. It owned the house he lived in and collected the rent. It owned the water and gas he consumed and billed him for them. It owned the building his grocer rented, so the grocer’s costs and therefore his prices were set by the company. It owned the arcade, the theatre, the library, the market hall, and the hotel. It owned the school building. It owned the church, which stood empty for long stretches because no congregation could afford the rent the company required of it. There were no independent newspapers, no public meetings, no political speeches, and no saloons, the only alcohol in town being served at the Hotel Florence bar for the benefit of visiting businessmen.
Company inspectors could enter houses. Residents held their leases at ten days’ notice and employment was, in practice, conditional on residence, so that losing a job meant losing a home and losing a home could mean losing a job. In 1887, when the surrounding township voted on annexation to Chicago, Pullman successfully petitioned to have his town excluded, preserving it from municipal government, water, and police that he did not control; the rest of the township joined the city in 1889 while Pullman remained outside. This is private government of a fairly complete kind, and the questions it raises about who may exercise authority over a population, and on what warrant, are exactly those examined in the record of privately governed jurisdictions, with the added feature of continuous observation familiar from any institution that monitors the people inside it. He owned the churches. He also owned the bar.
Ely’s Verdict
The most acute contemporary criticism arrived early, in February 1885, when the economist Richard T. Ely published a study of the town in Harper’s Monthly after visiting it. Ely was fair about the physical achievement and admiring of much of what he saw, and then reached a conclusion that has outlived every other verdict on the place: that the idea of Pullman was un-American, and that what the town amounted to was benevolent, well-wishing feudalism.
There was a second criticism that visitors made less often and that residents made constantly, which concerned the absence of anywhere to be unobserved. A town with no independent newspaper, no public meeting hall available for grievances, no saloon, and inspectors entitled to enter houses is a town in which organising is structurally difficult, and difficulty in coordinating is precisely what prevents a dispersed grievance from becoming a collective demand. Concealment and private communication are the preconditions for any coalition, in human institutions as in the natural world’s economy of signals and concealment, and Pullman had eliminated both. It worked for fourteen years, and then it failed all at once, which is the characteristic behaviour of suppressed rather than absent discontent.
The phrase is exact and it names the mechanism rather than the man. Ely was not accusing Pullman of cruelty, and there is little evidence of any; the complaint is that the arrangement made the residents’ welfare entirely dependent on the continuing goodwill of a proprietor whose interests they had no means of influencing, which is a description of a feudal relation regardless of how well the lord behaves. Pullman was stung by it and never really answered it, in part because the answer would have required him to give up the six per cent. It is worth noting how early the diagnosis arrived: nine years before the strike, a visiting economist identified the exact structural feature that would produce it, and nothing was done, which is the ordinary fate of accurate warnings delivered to institutions with a financial reason not to hear them, as the histories of organisations that could not act on their own audits repeatedly show. He got it right in 1885. Nobody had to wait for 1894.
One Counterparty
Here is the mechanism, stated generally. In an ordinary town, a person’s obligations are distributed among many parties with different interests: an employer, a landlord, a grocer, a utility, a bank, a church, a council. That dispersion is not designed and is rarely noticed, and it does something crucial. It means that a shock in one relationship can be absorbed by the others. Wages fall, and a landlord grants a month’s grace because an empty house earns nothing; a grocer extends credit because he wants the custom back; a parish fund helps; a council suspends a charge. Each of those parties acts from its own interest, and the aggregate effect is a shock absorber that nobody built.
There is a further effect worth naming, which concerns information rather than money. In a dispersed arrangement, a person’s various counterparties each know only part of their situation, and that partial ignorance is quietly protective: the grocer does not know what you earn, the landlord does not know whether you have been reprimanded at work, and none of them can act on the whole picture. Where every relationship runs to one party, that party knows everything, and the tenant knows that it knows. The transmission of practical knowledge about how to survive within such a system, which is ordinarily how communities protect their members, becomes correspondingly risky, and the informal networks through which people pass on what they have learned had nowhere safe to operate.
Pullman abolished it. Every one of those relationships ran to the same counterparty, which meant that when the shock came there was nothing to absorb it, because the party that might have granted relief in one relationship was the same party that had just imposed the loss in another, and it had already decided. It also meant something subtler and more corrosive: a wage cut and a rent bill were no longer separate events. On the company’s books the factory and the town were distinct accounts with distinct requirements, and cutting wages while holding rents was internally coherent bookkeeping. From the tenant’s side there was no distinction at all. The same organisation put less into one pocket and took the same amount out of the other, and because rent was deducted before the pay envelope was handed over, the arithmetic was performed in public, in front of the man it was performed on. Concentration of this kind removes every alternative and leaves grievance nowhere to go but into aggregate, which is how ordinary discontent in a bounded group becomes collective action, as the study of how coalitions form under pressure describes. One address received every bill. It also received every complaint.
1893
The Panic of 1893 produced the worst depression the United States had known. Orders for sleeping cars collapsed, and Pullman responded as a manufacturer reasonably might, by cutting wages between twenty and thirty per cent and spreading what work remained across as many hands as possible rather than laying men off outright. He did not reduce rents, because the town was a separate account with its own return requirement. He did not reduce prices at the company stores. He did not, by any account, offer any compensating adjustment at all.
The company’s defence of the wage cuts was not unreasonable on its own terms and deserves stating. Orders had genuinely collapsed, and Pullman took work at or below cost specifically to keep the shops open and the men employed rather than shutting down entirely, which he could have done and which many manufacturers did. Spreading reduced work across a full workforce at reduced rates is a defensible response to a depression and arguably a humane one. The objection is not that wages were cut. It is that the same organisation, having cut them on the manufacturing account for good reasons, then declined to adjust the rents on the property account, and that the two accounts belonged to one company facing one set of families.
The numbers from those two years are the argument in its most compressed form. In 1893 the company paid out roughly seven and a quarter million dollars in wages and about two and a half million in dividends. In 1894, after the cuts, it paid roughly four and a half million in wages and about two point nine million in dividends, and carried an undistributed surplus of over two million more. Wages fell by nearly three million dollars and dividends rose by several hundred thousand, which is a defensible outcome for a manufacturing company in a depression and an indefensible one for a landlord whose tenants are its own workforce. Whether those two roles could be held simultaneously is the whole question, and the company’s answer was that they could, because the accounts were separate. Some men that spring received pay envelopes containing a few cents after the rent deduction, and some received nothing. The books balanced. The envelope did not.
May 11, 1894
The sequence that followed was short and it is worth recording precisely, because the community behaved with considerable restraint and got nowhere. Workers formed a committee and on 7 May approached the company to ask that rents be reduced in line with wages. They were refused. On 9 May they asked again and were refused again. Three members of the committee were then dismissed, which the company denied was retaliation and which the workforce understood perfectly well. On 11 May roughly ninety per cent of the workforce failed to appear, and the company dismissed the remainder and closed the shops.
The rent arrears themselves tell a story that undercuts the company’s public position. Pullman argued that the town’s finances were entirely separate from the works and that rents could not be adjusted without breaching an obligation to shareholders, and yet during and after the strike the company made comparatively little effort to evict tenants or to collect the sums owed, which suggests the separation was rather less absolute than the argument required. An organisation that can decline to enforce a debt when enforcement would be embarrassing is an organisation that could have reduced the charge when reduction would have been useful, which is the kind of inconsistency that emerges whenever internal accounts are used to justify a position rather than to describe one, familiar from the forensic literature on how institutions represent their own books.
It is worth noting that this was not the first strike in Pullman, whatever the town’s reputation suggested; there had been unrest as early as 1882, within two years of the first residents arriving, which rather complicates the picture of a contented model community suddenly disturbed. The Illinois Labor History Society’s account of what began as a revolt against wage cuts and company practices traces how a local grievance escalated, and the escalation is the part that made history. Pullman refused arbitration, refused to meet the committee, and left for the east. He appears genuinely to have believed there was nothing to discuss, since the rents were set by the return requirement and the return requirement was not a matter of opinion. They asked twice. They were told no twice.
The Boycott
What turned a shutdown at one Illinois factory into a national emergency was the second consequence of concentration. Pullman did not merely make sleeping cars; Pullman cars ran on nearly every major railroad in the country under operating contracts. When the American Railway Union, led by Eugene V. Debs, voted in late June to support the strikers by refusing to handle any train carrying a Pullman car, the boycott did not target one company. It targeted the entire American rail network, because the entire American rail network had Pullman equipment attached to it.
The economic stakes explain the ferocity of the response. American commerce in 1894 moved by rail and by very little else, so a national rail stoppage was not an inconvenience but a suspension of the country’s ability to move grain, coal, cattle, manufactured goods, and mail, in the way that any interruption to a dominant transport artery halts the trade that depends on it, as the history of moving commodities to market makes clear. That is why the railroads combined, why the federal government intervened on a scale it had never previously attempted in a labour dispute, and why the mails provided the legal hook. It also explains why the union lost: a boycott powerful enough to stop a nation is powerful enough to guarantee that the nation will stop it.
Within weeks traffic was halted from Ohio to California, something between one hundred and fifty thousand and a quarter of a million workers were involved across twenty-seven states and territories, and the General Managers Association of the railroads had made common cause against the union. Attorney General Richard Olney, a railroad lawyer who continued to draw retainers while in office, obtained a sweeping federal injunction on the ground that the boycott obstructed the mails, and persuaded President Cleveland to send federal troops to Chicago on 3 July, over the objections of Governor Altgeld and the mayor. Rioting followed the soldiers rather than preceding them; railcars were burned, and roughly thirty people were killed. Debs was jailed for six months and read his way into socialism there. Cleveland, who had signed the act making Labor Day a federal holiday on 28 June, in the middle of all this, lost his party’s renomination two years later, which is the kind of political arithmetic that presidents discover after the fact. The strike was local. The company’s integration made it national.
Pullman in 2026
The legal ending is the most exact verdict anyone delivered. Pullman died in October 1897, sufficiently hated that he was buried at night in a pit lined with reinforced concrete and railway ties, for fear the grave would be violated. His successor as company president was Robert Todd Lincoln. In 1898 the Illinois Supreme Court, on quo warranto proceedings brought by the state, ordered the company to divest itself of all property not used for manufacturing, finding that its charter authorised it to make railway cars and said nothing whatever about owning a town, and observing that company towns were opposed to good public policy. The company took until 1907 to comply. The houses were sold, largely to their occupants, and Pullman became an ordinary Chicago neighbourhood.
It is still there and it is largely intact, which makes it the rare entry in this territory that is neither ruin nor archive. The Hotel Florence, the Arcade site, the Greenstone Church, the administration building with its clock tower, and street after street of the original brick housing survive, occupied, in a district designated a national monument in 2015 and redesignated Pullman National Historical Park in 2022, with a visitor centre in the former factory administration building. The Pullman porters, the Black men who staffed the sleeping cars for tips and who organised the Brotherhood of Sleeping Car Porters under A. Philip Randolph, are commemorated there too, and their history is arguably the more consequential legacy. The neighbourhood is a functioning part of Chicago rather than a place preserved only in the record, documented now with the survey and monitoring tools, including aerial and remote sensing equipment, applied to any historic district. The town outlived the arrangement. That was the point of the ruling.
The Address
Strip Pullman to its structure and the paradox dissolves. This was simultaneously the best worker housing in America and the origin of the century’s most violent labour conflict, and those are not two facts requiring reconciliation. They are one fact. The quality of the houses was never the issue and improving them further would have changed nothing, because the problem was not what the company provided but that the company provided all of it. Concentrating employment, tenancy, utilities, retail, worship, education, and government into a single counterparty removes the diversification that lets ordinary economic shocks be absorbed by somebody, and it converts every grievance into a grievance against the same party, which is the definition of a general dispute. Then, because the company was equally concentrated in its market, the general dispute acquired a national lever.
The test this suggests is worth applying wherever an institution provides more than one thing to the same people. Ask how many distinct relationships a person has with the organisation, and what happens to all of them simultaneously if any one goes wrong. Ask whether the accounts that justify a decision on one side are visible to the people experiencing it on the other. Ask who could grant relief in a downturn, and whether that party is the same party imposing the loss. And ask what would absorb an ordinary shock, because in a concentrated arrangement the honest answer is nothing at all.
Pullman himself seems never to have understood the objection, and there is something almost poignant in that. He had built exactly what he said he would build, to a standard nobody disputed, and he had been transparent from the beginning that it was required to earn its keep. What he had not seen was that the arrangement which made the town possible also made it impossible to be a good landlord and a hard-pressed employer at the same time to the same people, and that the goodwill on which the whole thing rested was the one asset he had no way to book. That is what earns the case its place in the catalogue of utopian societies, among the paternalisms rather than the communes, and among the failures whose mechanism is fully legible. He built a model town and required it to yield six per cent, and in the summer of 1894 he discovered what the other four per cent had been paying for.